Security Analysis About The Train Control Center Based On Bayesian Networks

With the development of rail transport, the role of high-speed railway is more important in promoting economic development and social progress. We planned and constructed four vertical and four horizontal high-speed rail networks, the train speed continues to improve, train traffic density increases, which has higher requirements on traffic safety, reliability. Train control system is the core equipment of high-speed trains which is real-time control, and train control center is an important part of the ground train control system, it plays an important role in ensuring the high-speed railway safety and improving transport efficiency. Train control center combines computer technology, communication technology, automatic control technology, and it is complex, powerful, so it has a higher demand for safety. Therefore, from the point of the view of safety system engineering, it is a very valuable research that we use qualitative and quantitative research methods to evaluate the various factors affecting the high-speed railway safety, in order to establish appropriate safety measures.We will keep focus on the core equipment-the train control center of the train control system in this paper. On the basis of the relevant technical specifications train control system overall technical program, the system requirements specification, functional requirements specification etc. we will study the analysis methods of the safety of train control center, and use Bayesian networks to analyze the safety of the train control center. The main contents and results of the paper include:classify the hierarchical structure of the train control center, construct the train control center architecture reference model to describe the system from the perspective of the composition of the train control center system composed of static structure, to determine its boundaries; analyze the function of the train control center and refinement, and construct the hierarchical model of functional architecture reference model, describe the main functions and sub-functions of the Train Control Center from the perspective of functional level; analysis of the train control center function execution process, construct the train control center function sequence diagram model, dynamically describe the train control center’s every function of the specific implementation process and the interaction between the various object.On this basis, construct Bayesian network for the train control center as a whole from the structure, analysis of the actual use of equipment how to affect the train control center. Aiming at priori probability of the root node in the Bayesian network model,build a Bayesian network algorithm based on multi-expert judgment to get the root of a priori probabilities, and then use Bayesian network inference algorithm based on multi-tree algorithm and clustering algorithm to do causal reasoning and diagnostic analysis, quantitatively calculate the posterior probability of normal and faults of the train control center, possible causes of the failure occurred, and the posterior probability analysis of various factors. Finally, build a Bayesian network model to the main function of the train control center, do the qualitative and quantitative analysis of the various nodes during the execution of a function to the influence of weight function. For example, the function in track circuit code, analyze the structure and function of the train control center to get the node of the Bayesian network model. According to the function of the track circuit code execution, build a Bayesian network model based on function of track circuit code, quantitative calculate the contribution of various nodes and weights as the track circuit code function is in different state, and do the analysis of the coding implementation track circuit fault diagnosis.